This is a prospective study of 4800 consecutive admissions to the Psychiatric Service of the Houston VA Hospital. On each patient certain ratings were made at time of admission and recorded in standardized form. The 4800 subjects are being followed up for a period of two to four years. There names are being checked against lists of death certificates for instances of death by suicide, homicide, accident, and cirrhosis. The names are also checked against lists of readmissions to the Houston VA Hospital, particularly if for suicide attempt or suicide ideation, but also for admissions for alcoholism, drug abuse and other types of "self-destructive" behavior. During the first project year, several months were devoted to pilot trials for reliability checks and refinements of rating instruments. The patient intake period began January 1, 1972, and was completed on January 31, 1974, at which time 4800 separate patients had been enrolled. Of this group, 810 or 16.9 percent were identified as "high risk" subjects by criteria described in the original protocol. This "high risk" group and a ramdomly selected control group of 743 subjects have both been given extensive additional interviews, ratings, and questionnaires. To date there have been 26 completed suicides. Follow-up of the total group is continuing.